Six specific statistical research areas have been identified which are of a general nature and relate to issues in the data analysis of clinical and laboratory research from the Wisconsin Clinical Cancer Center or from cancer research in general. 1. Analysis of Censored Survival Data a) Nonparametric Regression Analysis b) Robust Estimation for Linear Models c) Discrete Time Survival Estimates with Time-Dependent Covariates 2. Clinical Trial Data Monitoring a) Group Sequential Methods b) Stochastic Curtailed Sampling 3. Goodness of Fit a) Asymptotic Properties for a Large Class of Tests b) Minimum Distance Estimation for Multinomial Probabilities c) Proportional Hazard Models 4. Selected Sampling in Regression Analysis 5. Stratified Randomization vs. Non-Stratified Randomization 6. Other Research Problems The investigators will interact with each other in this research but individuals will be primarily responsible for various aspects. Statistical methodology will be used including large sample theory, normal theory, numerical integration, Monte Carlo methods, and graphical techniques using the Department of Statistics' VAX/750 computer. Objectives are the development of more appropriate or improved methods for the analysis of data in the general medical research setting with exportable computer software as needed.